268
Biochimica 0 Elsevier
et Biophysics
Scientific
Acta,
Publishing
441
(1976)
Company,
268-279
Amsterdam
-Printed
in The Netherlands
BBA 56832
CHOLESTEROL
SULFATE
IN RAT TISSUES
TISSUE DISTRIBUTION, DEVELOPMENTAL SUBCELLULAR LOCALIZATION MASAO
IWAMORI,
HUGO W. MOSER
and YASUO
Eunice Kennedy Shriver Center for Mental Retardation Waltham, Mass. 02154 and Department of Neurology, Boston, Mass. 02114 (U.S.A.)
(Received
January
27th,
CHANGE AND BRAIN
KISHIMOTO at Walter E. Fernald State School, Massachusetts General Hospital,
1976)
Summary 1. A reliable micromethod for the determination of the tissue level of cholesterol sulfate has been developed. Cholesterol sulfate was separated from the bulk of the free cholesterol by silica gel column chromatography, and the cholesterol sulfate fraction subjected to benzoylation. A small amount of contaminating free cholesterol and other lipids remaining in this fraction were converted to benzoyl esters while the cholesterol sulfate remained unreacted. The cholesterol sulfate was then separated from the benzoylated contaminants by a second silica gel chromatography column and subjected to solvolysis. The liberated cholesterol was determined by gas-liquid chromatography. 2. The cholesterol sulfate contents of the visceral organs of 43-day-old rats were determined. Every tissue examined contained small amounts of this sulfate. Kidney contained the highest concentration of cholesterol sulfate (250300 pg/g dry tissue weight) followed by spleen (77 lg/g), adrenal gland (50-70 pglg) and lung (50-57 pug/g). 3. In brain, cholesterol sulfate level rises sharply from 17 pg/g dry weight in 7day-old rats to more than 50 pg/g in 15-day-olds, then it declines rapidly to 15 pg/g in the 40-day-olds and this level is maintained to adulthood. The developmental pattern in the liver resembles that in the brain, except that the peak is somewhat flatter with the highest value (60 pg/g dry weight) occurring in the 21day-old animal. In contrast to the above two tissues, the level of kidney cholesterol sulfate increases steadily from 15 E.cg/gin 7-day-olds and reaches the adult level of approx. 200 Erg/g in 50-day-olds. 4. The highest level of cholesterol sulfate in subcellular fractions of rat brain occurred in a fraction rich in nerve endings. The level here was 10 times higher than that in the mitochondrial fraction, which contained the lowest levels of this steroid sulfate.
269
Introduction Since Drayer et al. [l] first isolated cholesterol sulfate from bovine adrenal gland in 1964, its occurrence has been reported in many invertebrate [2--41 and mammalian tissues [5-lo]. The human tissues studied thus far include: brain, liver, kidney, plasma, urine, bile, and feces, which were examined earlier in this laboratory [5] and blood, gallstones, aortic placques, and normal and malignant adrenal tissue, as reported by Drayer and Lieberman [7,8]. Coles et al. [9] also reported the tentative identification of cholesterol sulfate in mouse kidney and liver. More recently, Bleau et al. [lo] established a sophisticated assay for cholesterol sulfate by enzymatic radioisotopic displacement and determined sulfate levels in plasma, erythrocytes, semen, and saliva, all from human tissue. The presence of abnormally elevated cholesterol and cerebroside sulfates has been observed in the liver [ll], kidney, plasma and urine [ 121 of a patient with metachromatic leukodystrophy with multiple sulfatase deficiency. Several reports appeared in the literature which suggest cholesterol sulfate has a physiological role in these tissues. For example, cholesterol sulfate is converted into several sulfate esters of steroid hormones [13,14] and it contributes toward the stability of erythrocyte membrane [ 15,161. Despite the increasing evidence of the significance of cholesterol sulfate in biological systems, little is known about the nature and quantity of this material in mammalian visceral organs. We report here a reliable microprocedure which is suitable for measuring the minute amounts of cholesterol esters in various rat tissues. In addition, we have studied the changes of cholesterol sulfate levels in brain, liver, ard kidney during maturation and, further, its distribution in subcellular fractions of rat brain. Materials and Methods Materials. The sodium salt of cholesterol sulfate was synthesized in this laboratory [ 51. [ 4-14C] cholesterol (57.5 Ci/mol), [ 1 ,2-3H] cholesterol sulfate (46.4 Ci/mmol) and [7-3H]dehydroepiandrosterone sulfate (25.1 Ci/mmol) were purchased from New England Nuclear, Boston, Mass. Benzoyl chloride was obtained from Eastman Chemicals, Rochester, N.Y. NADH and 2-@-iodophenyl)3-@-nitrophenyl)-5-phenyltetrazolium were obtained from P. L. Biochemicals, Inc., Milwaukee, Wise. and British Drug House, Poole, England, respectively. DMannitol, obtained from Fischer Scientific, Pittsburg, Pa., was recrystallized from ethanol/water (1 : 1, v/v) and cholestane, purchased from Supelco Inc., Bellefonte, Pa. was recrystallized from benzene/ether (4 : 1, v/v). Cerebroside sulfates were prepared from human brain by a published procedure [ 171. All solvents were redistilled before use. Pyridine was redistilled over KOH and stored over KOH and molecular sieve 5A (Fisher Scientific). Non-radioactive or [ l-14C] cholesterol Synthesis of cholesterol benzoate. (57.5 Ci/mol) was heated with an excess amount of benzoyl chloride in dry pyridine for 30 min at 60°C. To convert unreacted benzoyl chloride to methyl benzoate, methanol was added to the reaction mixture and again heated at
27@
60°C for 30 min. The reaction mixture was evaporated to dryness under a flow of N2. The residue was dissolved in chloroform and washed with 0.05 M Na$O, and then with methanol/water (1 : 1, v/v) until neutrality was obtained. After removal of the solvent the final product was purified by preparative thin-layer chromatography on a silica gel GF plate (0.25 mm thick, Analtech, Newark, Del.) with hexane/ether (4 : 1, v/v) as a developing solvent. The band of cholesterol benzoate was located by ultraviolet light and eluted with ether. Determination of cholesterol sulfate. Sprague-Dawley rats were sacrificed by decapitation and the tissues were immediately removed and weighed. The tissues were homogenized with small amounts of water and lyophilized. The dried residues were extracted three times with chloroform/methanol (2 : 1, v/v) (9 ml/g of fresh tissue) at 45°C and the combined extracts were washed according to Folch et al. [ 181. The lower organic phase was evaporated to dryness in vacua and the residue was subjected to mild alkaline methanolysis [19]. The lipoidal product of methanolysis was dissolved in a small volume of chloroform/methanol (97 : 3, v/v), applied to a column of Unisil (a silica gel, 100200 mesh, Clarkson Chemical Co., Williamsport, Pa.; 1 g of Unisil per 30 mg lipid) and washed with the same solvent (20 ml/g Unisil). The column was then eluted with chloroform/methanol (85 : 15, v/v) (40 ml/g Unisil). The latter solvent eluted cholesterol sulfate, cerebrosides, sulfatides and a small amount of free cholesterol. In order to remove the free cholesterol, the eluate was treated with benzoyl chloride as described above with the following modification: the eluate was dissolved in 0.45 ml of dry pyridine and 50 ~1 of benzoyl chloride was added; the mixture was heated for 20 min at 37°C. Excess benzoyl chloride was removed as described above. After washing with 0.05 M Na,C03 the product was fractionated on a Unisil column as described above. The remaining free cholesterol, cerebrosides, and sulfatides were converted into nonpolar benzoates and eluted with the first solvent, chloroform/methanol (97 : 3, v/v). Recovery studies referred below showed that this conversion was quantitative. Cholesterol sulfate was unaffected by this treatment and was collected by a second elution with chloroform/methanol (85 : 15, v/v). The cholesterol sulfate fraction was hydrolyzed by dissolving it in 0.5 ml of chloroform/methanol (2 : 1, v/v) and adding 15 1.11of 1 M methanolic HCl. The mixture was allowed to stand at 37°C for 15 h and was then neutralized with 1 ml of NazC03-saturated methanol. 10 ~1 of internal standard (0.5 mg cholestane/ml benzene) 2 ml chloroform, and 1 ml water were added to the neutralized material and well mixed. The lower chloroform phase, containing desulfated free cholesterol, was evaporated to dryness under Nz and the residue was further dried over P,05 in an evacuated desiccator. The dried residue was heated with N,O-bis-( trimethylsilyl)-acetamide/trimethylchlorosilane (5 : 1, v/v) at 80°C for 20 min and the reaction mixture was analyzed by gas-liquid chromatography for cholesterol content. The sample was analyzed by a Hewlett-Packard Model 7620A gas chromatograph equipped with flame ionization detectors and a column containing 3% OV-1 coated on Chromosorb W, (SO-100 mesh, AW-DMCS, Applied Science Lab., State College, Pa.) at an oven temperature of 215°C. The molar equivalent of cholesterol sulfate was calculated by comparing the peak area of cholesterol with that of the internal standard (cholestane) by the
271
following equation: nmol of cholesterol
sulfate
= nmol of cholestane
added
x
area of cholesterol area of cholestane
x
1.0789
Peak areas were determined by the cut-and-weigh method. The last number, a response factor, was derived from peak areas obtained from a mixture of known amounts of cholesterol and cholestane. The free cholesterol determination was performed by an identical gas-liquid chromatographic procedure. The chloroform/methanol (97 : 3, v/v) eluate of the first Unisil column chromatography fractionation of the total tissue lipid extract was used as the material for this dete~ination. ~u~ce~lul~r fractionation of bruin. Nerve endings and mitochond~a fractions were prepared from ten whole brains of 30-day-old male rats by the method of DeRobertis and his colleagues [20,21] as modified by Clendenon and Allen 1221. Other analytical procedures. Protein was determined by the method of Lowry et al. [ 231 using bovine serum albumin (Miles Laboratories, Inc.) as the standard. Total lipid phosphorus was measured according to Bartlett [ 241 after wet digestion of lipids with HC104 and H202. Total lipid galactose was measured as its trimethylsilyl derivative by gas-liquid chromatography [25]; D-mannitol was used as the internal standard. Infrared spectra were obtained from a KBr pellet with Beckman IR-33 spe~trophotometer. Gas-liquid chromatography-mass spec~omet~ was performed as described previously 1263. Electron micrographs were obtained with Philips 300 electron microscope operating at 60 kV. The brain subcellular fractions were treated by the procedure of Swanson et al. [ 271. Succinate dehydrogenase and lactate dehydrogenase were assayed by the procedures of Pennington [28] and Komberg [29], respectively. Results Validation of cholesterol sulfate determination The present procedure determines the free cholesterol liberated after the solvolysis of cholesterol sulfate. Since rat tissues contain very small amounts of this sulfate and much larger amounts of free cholesterol, complete removal of free cholesterol from the cholesterol sulfate fraction was an absolute requirement. To determine the effectiveness of Unisil chromatography for the removal of free cholesterol, 5480 cpm of [4-‘4C]cholesterol was added to 5-10 mg of the alkali-treated total lipids of either brain or liver and the lipids were fractionated as described above. Although most of the free cholesterol was removed by the first elution with chloroform/methanol (97 : 3, v/v) a small amount of radioactivity (0.26-3.64%) was eluted with the second solvent of chloroform/methanol (85 : 15, v/v). To remove the contaminating cholesterol and sphingolipids, which came in this fraction, we converted them to benzoyl esters and removed with a second Unisil chromatography step. The colesterol sulfate fraction obtained from this second ~hromato~aphy contained no radioactivity. Approx. 90% of cholesterol was benzoylated after 20 min reaction as described above. Higher temperature accelerated the benzoylation, but partial cleavage of the sulfate bond occurred; 17% of the cholesterol sulfate was cleav-
272
ed after 30 min heating at 80°C. The solvolysis of cholesterol sulfate, cerebroside sulfate and dehydroepiandrosterone sulfate also was completed in 4 h. The recovery of cholesterol sulfate in this procedure was examined by adding [l,2-3H]cholesterol sulfate (98 937 cpm) to 12 mg of brain total lipids. The mixture was treated as described. The radioactivity recovered in the final hydrolyzed product was 88.4%. Recovery of the added cholesterol sulfate in the final product was constant between 1 and 100 pg (Fig. 1; part of the line is not shown). The cholesterol sulfate obtained with this procedure was identified by several methods. Cholesterol sulfate was isolated from 450 g of calf adrenal by the identical procedure. Thin-layer chromatography (silica gel G as absorbent and n-propanol/concentrated ammonia/water (12 : 2 : 1, v/v as solvent) of the fraction obtained from the second Unisil column revealed a largespot with a mobility identical to that of authentic sulfate and a few minor contaminating spots. This major component was further purified by preparative thin-layer chromatography under the same conditions as described above. The band was located by spraying with bromothymol blue reagent and eluted with chloroform/ methanol/water (20 : 10 : 1, v/v). The solution was washed with water [18] until it was colorless; evaporation of the solvent yielded 5.0 mg of residue. This residue showed the same Rf value on thin-layer chromato~ams as synthetic cholesterol sulfate. The residue was converted to a sodium salt by the use of a Dowex 50.AGX-8 column [ 301. The melting point of this salt was 193-195”C,
24
11
0
Cholesteryl
Sulfate
Added
nmol
Fig. 1. Recovery of added cholesterol sulfate as free cholesterol by the present procedure. Details of the experiment were given in text.
273
while that of synthetic cholesterol sulfate sodium salt was 194-196°C: the mixed melting point was 194-197°C. These values agreed well with published data [31]. The infrared spect~m of this sulfate was identical to that of the authentic sample as previously reported f5]. The lipid portion of this sulfate was identified as cholesterol by thin-layer chromatography and gas-liquid chromatography. For the latter procedure, the material obtained from solvolysis of the cholesterol sulfate was converted to the trimethylsilyl derivative and analyzed on both OV-1 and DEGS columns. Only a single peak corresponding to cholesterol t~methylsilyl ether was obtained in both chromato~~s. Further identification of this ether was performed by gas-liquid chromatography-mass spectrometry: the molecular ion at n/e 458, as well as other characteristic fragments, such as m/e 368, 353, 329 and 255 [32] were unequivocally obtained. Concentration of ~h~~estero~ sulfate in cursors rat tissues The sulfate was present in all tissues examined (Table I). The highest concentration of the sulfate was obtained in the kidney, followed by spleen and adrenal gland. There was no significant difference due to the sex of the animal. Developmental change of cholesterol sulfate. Figs. 2A, B and C illustrate the change of tissue levels of cholesterol sulfate and of free cholesterol in rat brain,
TABLE
I
CONCENTRATIONS Data
were
presented
OF
CHOLESTEROL
as mean
Tissue
values
Tissue
obtained dry
SULFATE from
weight
IN
two
VARIOUS
separate Cholesterol dry
TISSUES
OF
43-DAY-OLD-RATS
experiments. sulfate weight)
Cholesterol (mg/g
19)
Wglg
Liver
3.72
16.40
Thymus
0.15
4.19
27.68
Lung
0.34
49.52
39.75
Heart
0.29
9.77
14.72
Spleen
0.18
77.45
33.61
249.37
20.40
Male
Kidney
0.46
Brain
0.17
20.94
Adrenal
0.01
69.38
Eye
0.03
Testis
0.27
Prostate
0.07
8.25
87.40
1.77
*
66.39
*
13.78
34.06
28.10
3.59
*
5.91
Female Liver
2.83
Thymus
0.11
4.49
24.77 41.41
20.22
8.16
Lung
0.26
56.95
Heart
0.13
9.66
15.46
Spleen
0.15
76.48
32.92
Kidney
0.37
Brain
0.72
16.24
Adrenal
0.02
50.80
Eye
0.03
2.03
*
12.92
uterus
0.05
11.51
*
31.51
Ovaries
0.02
34.31
*
52.05
* These
values
were
obtained
306.55
by pooling
tissues
from
15.11
10
71.97 *
animaIs
67.55
of the same
age,
dry
weight)
274
, /,’
x’
,r” \
1 ,L
1.\..
.
L
-ti
L
0
.
, IO
1
20
30 Age
10
50
days
Fig. 2. Change of amounts of cholesterol sulfate (open circles) and free cholesterol (solid circles) as a function of age in rat brain (A), liver (B) and kidney (C). The methods of assay were described in text. Each circle represents single determination.
275
liver, and kidney, respectively. The cholesterol sulfate level in the brain rose sharply after the 7th postnatal day, peaked at 15 days, then declined to reach the adult level at 50 days, a level which is below that observed at 7 days. This is in contrast with the free cholesterol level which increases steadily until it attains a plateau at 50 days. Liver followed an analogous developmental pattern for cholesterol sulfate; however, the peak was much broader than that of brain and the highest level was obtained at 20-25 days. The free cholesterol level in liver remained almost constant throughout the period examined. In contrast with the above two tissues, the level of cholesterol sulfate in kidney increased steadily from day 7 to day 60. After that the level plateaued. The level of free cholesterol in kidney remained constant. The rate of deposition of cholesterol sulfate in brain, liver, and kidney in lo-day-old rat was calculated as 3.75, 17.33, and 4.26 pg/g of dry tissue weight per day, respectively. Subcellular distribution of cholesterol sulfate in brain Table II shows concentrations of cholesterol sulfate in subcellular fractions of rat brain. Distributions of protein, phospholipids, cerebrosides and free cholesterol were also shown in the table, and are in good agreement with those previously reported [ 33-361. The highest concentration of cholesterol sulfate was observed in the nerve ending fractions, the level being 2.5 times higher than whole homogenate and 10 times higher than the mitochondrial fraction. The nerve ending fractions were shown to be reasonably pure by electron microscopy (Fig. 3) and marker enzyme activities: lactate dehydrogenase for nerve endings and succinate dehydrogenase for mitochondria (also shown in Table II).
TABLE
II
SUBCELLULAR
DISTRIBUTION
OF CHOLESTEROL
SULFATE
IN 3-DAY-OLD
RAT BRAIN
Values are the average of three separate experiments. See text for the method of subcellular fractionation. Mitochondria I and nerve endings I were obtained from the first gradient centrifugation and mitochondria II and nerve endings II were obtained from mitochondria I by the second gradient centrifugation. Fractions
Protein (mgl brain)
Cholester01 sulfate (nmol/ mg protein)
Homogenate Nuclei cell debris Myelin Mitochondria I Mitochondria II Nerve endings I Nerve dneings II Microsomes CYtosol n.d., not determined.
133.1 27.3 14.5 43.9 14.3 6.4 11.8 15.3 23.0
0.04 0.04 0.05 0.02 0.01 0.10 0.12 0.02 0.00
Cholester01
Cerebroside
Phosphorus
Wmoll mg protein)
Wmol/ mg protein)
WmoI/ mg protein)
0.29 0.28 0.62 0.30 0.10 0.57 0.58 0.27
Biochimica 0 Elsevier
et Biophysics
Scientific
Acta,
Publishing
441
(1976)
Company,
268-279
Amsterdam
-Printed
in The Netherlands
BBA 56832
CHOLESTEROL
SULFATE
IN RAT TISSUES
TISSUE DISTRIBUTION, DEVELOPMENTAL SUBCELLULAR LOCALIZATION MASAO
IWAMORI,
HUGO W. MOSER
and YASUO
Eunice Kennedy Shriver Center for Mental Retardation Waltham, Mass. 02154 and Department of Neurology, Boston, Mass. 02114 (U.S.A.)
(Received
January
27th,
CHANGE AND BRAIN
KISHIMOTO at Walter E. Fernald State School, Massachusetts General Hospital,
1976)
Summary 1. A reliable micromethod for the determination of the tissue level of cholesterol sulfate has been developed. Cholesterol sulfate was separated from the bulk of the free cholesterol by silica gel column chromatography, and the cholesterol sulfate fraction subjected to benzoylation. A small amount of contaminating free cholesterol and other lipids remaining in this fraction were converted to benzoyl esters while the cholesterol sulfate remained unreacted. The cholesterol sulfate was then separated from the benzoylated contaminants by a second silica gel chromatography column and subjected to solvolysis. The liberated cholesterol was determined by gas-liquid chromatography. 2. The cholesterol sulfate contents of the visceral organs of 43-day-old rats were determined. Every tissue examined contained small amounts of this sulfate. Kidney contained the highest concentration of cholesterol sulfate (250300 pg/g dry tissue weight) followed by spleen (77 lg/g), adrenal gland (50-70 pglg) and lung (50-57 pug/g). 3. In brain, cholesterol sulfate level rises sharply from 17 pg/g dry weight in 7day-old rats to more than 50 pg/g in 15-day-olds, then it declines rapidly to 15 pg/g in the 40-day-olds and this level is maintained to adulthood. The developmental pattern in the liver resembles that in the brain, except that the peak is somewhat flatter with the highest value (60 pg/g dry weight) occurring in the 21day-old animal. In contrast to the above two tissues, the level of kidney cholesterol sulfate increases steadily from 15 E.cg/gin 7-day-olds and reaches the adult level of approx. 200 Erg/g in 50-day-olds. 4. The highest level of cholesterol sulfate in subcellular fractions of rat brain occurred in a fraction rich in nerve endings. The level here was 10 times higher than that in the mitochondrial fraction, which contained the lowest levels of this steroid sulfate.
269
Introduction Since Drayer et al. [l] first isolated cholesterol sulfate from bovine adrenal gland in 1964, its occurrence has been reported in many invertebrate [2--41 and mammalian tissues [5-lo]. The human tissues studied thus far include: brain, liver, kidney, plasma, urine, bile, and feces, which were examined earlier in this laboratory [5] and blood, gallstones, aortic placques, and normal and malignant adrenal tissue, as reported by Drayer and Lieberman [7,8]. Coles et al. [9] also reported the tentative identification of cholesterol sulfate in mouse kidney and liver. More recently, Bleau et al. [lo] established a sophisticated assay for cholesterol sulfate by enzymatic radioisotopic displacement and determined sulfate levels in plasma, erythrocytes, semen, and saliva, all from human tissue. The presence of abnormally elevated cholesterol and cerebroside sulfates has been observed in the liver [ll], kidney, plasma and urine [ 121 of a patient with metachromatic leukodystrophy with multiple sulfatase deficiency. Several reports appeared in the literature which suggest cholesterol sulfate has a physiological role in these tissues. For example, cholesterol sulfate is converted into several sulfate esters of steroid hormones [13,14] and it contributes toward the stability of erythrocyte membrane [ 15,161. Despite the increasing evidence of the significance of cholesterol sulfate in biological systems, little is known about the nature and quantity of this material in mammalian visceral organs. We report here a reliable microprocedure which is suitable for measuring the minute amounts of cholesterol esters in various rat tissues. In addition, we have studied the changes of cholesterol sulfate levels in brain, liver, ard kidney during maturation and, further, its distribution in subcellular fractions of rat brain. Materials and Methods Materials. The sodium salt of cholesterol sulfate was synthesized in this laboratory [ 51. [ 4-14C] cholesterol (57.5 Ci/mol), [ 1 ,2-3H] cholesterol sulfate (46.4 Ci/mmol) and [7-3H]dehydroepiandrosterone sulfate (25.1 Ci/mmol) were purchased from New England Nuclear, Boston, Mass. Benzoyl chloride was obtained from Eastman Chemicals, Rochester, N.Y. NADH and 2-@-iodophenyl)3-@-nitrophenyl)-5-phenyltetrazolium were obtained from P. L. Biochemicals, Inc., Milwaukee, Wise. and British Drug House, Poole, England, respectively. DMannitol, obtained from Fischer Scientific, Pittsburg, Pa., was recrystallized from ethanol/water (1 : 1, v/v) and cholestane, purchased from Supelco Inc., Bellefonte, Pa. was recrystallized from benzene/ether (4 : 1, v/v). Cerebroside sulfates were prepared from human brain by a published procedure [ 171. All solvents were redistilled before use. Pyridine was redistilled over KOH and stored over KOH and molecular sieve 5A (Fisher Scientific). Non-radioactive or [ l-14C] cholesterol Synthesis of cholesterol benzoate. (57.5 Ci/mol) was heated with an excess amount of benzoyl chloride in dry pyridine for 30 min at 60°C. To convert unreacted benzoyl chloride to methyl benzoate, methanol was added to the reaction mixture and again heated at
27@
60°C for 30 min. The reaction mixture was evaporated to dryness under a flow of N2. The residue was dissolved in chloroform and washed with 0.05 M Na$O, and then with methanol/water (1 : 1, v/v) until neutrality was obtained. After removal of the solvent the final product was purified by preparative thin-layer chromatography on a silica gel GF plate (0.25 mm thick, Analtech, Newark, Del.) with hexane/ether (4 : 1, v/v) as a developing solvent. The band of cholesterol benzoate was located by ultraviolet light and eluted with ether. Determination of cholesterol sulfate. Sprague-Dawley rats were sacrificed by decapitation and the tissues were immediately removed and weighed. The tissues were homogenized with small amounts of water and lyophilized. The dried residues were extracted three times with chloroform/methanol (2 : 1, v/v) (9 ml/g of fresh tissue) at 45°C and the combined extracts were washed according to Folch et al. [ 181. The lower organic phase was evaporated to dryness in vacua and the residue was subjected to mild alkaline methanolysis [19]. The lipoidal product of methanolysis was dissolved in a small volume of chloroform/methanol (97 : 3, v/v), applied to a column of Unisil (a silica gel, 100200 mesh, Clarkson Chemical Co., Williamsport, Pa.; 1 g of Unisil per 30 mg lipid) and washed with the same solvent (20 ml/g Unisil). The column was then eluted with chloroform/methanol (85 : 15, v/v) (40 ml/g Unisil). The latter solvent eluted cholesterol sulfate, cerebrosides, sulfatides and a small amount of free cholesterol. In order to remove the free cholesterol, the eluate was treated with benzoyl chloride as described above with the following modification: the eluate was dissolved in 0.45 ml of dry pyridine and 50 ~1 of benzoyl chloride was added; the mixture was heated for 20 min at 37°C. Excess benzoyl chloride was removed as described above. After washing with 0.05 M Na,C03 the product was fractionated on a Unisil column as described above. The remaining free cholesterol, cerebrosides, and sulfatides were converted into nonpolar benzoates and eluted with the first solvent, chloroform/methanol (97 : 3, v/v). Recovery studies referred below showed that this conversion was quantitative. Cholesterol sulfate was unaffected by this treatment and was collected by a second elution with chloroform/methanol (85 : 15, v/v). The cholesterol sulfate fraction was hydrolyzed by dissolving it in 0.5 ml of chloroform/methanol (2 : 1, v/v) and adding 15 1.11of 1 M methanolic HCl. The mixture was allowed to stand at 37°C for 15 h and was then neutralized with 1 ml of NazC03-saturated methanol. 10 ~1 of internal standard (0.5 mg cholestane/ml benzene) 2 ml chloroform, and 1 ml water were added to the neutralized material and well mixed. The lower chloroform phase, containing desulfated free cholesterol, was evaporated to dryness under Nz and the residue was further dried over P,05 in an evacuated desiccator. The dried residue was heated with N,O-bis-( trimethylsilyl)-acetamide/trimethylchlorosilane (5 : 1, v/v) at 80°C for 20 min and the reaction mixture was analyzed by gas-liquid chromatography for cholesterol content. The sample was analyzed by a Hewlett-Packard Model 7620A gas chromatograph equipped with flame ionization detectors and a column containing 3% OV-1 coated on Chromosorb W, (SO-100 mesh, AW-DMCS, Applied Science Lab., State College, Pa.) at an oven temperature of 215°C. The molar equivalent of cholesterol sulfate was calculated by comparing the peak area of cholesterol with that of the internal standard (cholestane) by the
271
following equation: nmol of cholesterol
sulfate
= nmol of cholestane
added
x
area of cholesterol area of cholestane
x
1.0789
Peak areas were determined by the cut-and-weigh method. The last number, a response factor, was derived from peak areas obtained from a mixture of known amounts of cholesterol and cholestane. The free cholesterol determination was performed by an identical gas-liquid chromatographic procedure. The chloroform/methanol (97 : 3, v/v) eluate of the first Unisil column chromatography fractionation of the total tissue lipid extract was used as the material for this dete~ination. ~u~ce~lul~r fractionation of bruin. Nerve endings and mitochond~a fractions were prepared from ten whole brains of 30-day-old male rats by the method of DeRobertis and his colleagues [20,21] as modified by Clendenon and Allen 1221. Other analytical procedures. Protein was determined by the method of Lowry et al. [ 231 using bovine serum albumin (Miles Laboratories, Inc.) as the standard. Total lipid phosphorus was measured according to Bartlett [ 241 after wet digestion of lipids with HC104 and H202. Total lipid galactose was measured as its trimethylsilyl derivative by gas-liquid chromatography [25]; D-mannitol was used as the internal standard. Infrared spectra were obtained from a KBr pellet with Beckman IR-33 spe~trophotometer. Gas-liquid chromatography-mass spec~omet~ was performed as described previously 1263. Electron micrographs were obtained with Philips 300 electron microscope operating at 60 kV. The brain subcellular fractions were treated by the procedure of Swanson et al. [ 271. Succinate dehydrogenase and lactate dehydrogenase were assayed by the procedures of Pennington [28] and Komberg [29], respectively. Results Validation of cholesterol sulfate determination The present procedure determines the free cholesterol liberated after the solvolysis of cholesterol sulfate. Since rat tissues contain very small amounts of this sulfate and much larger amounts of free cholesterol, complete removal of free cholesterol from the cholesterol sulfate fraction was an absolute requirement. To determine the effectiveness of Unisil chromatography for the removal of free cholesterol, 5480 cpm of [4-‘4C]cholesterol was added to 5-10 mg of the alkali-treated total lipids of either brain or liver and the lipids were fractionated as described above. Although most of the free cholesterol was removed by the first elution with chloroform/methanol (97 : 3, v/v) a small amount of radioactivity (0.26-3.64%) was eluted with the second solvent of chloroform/methanol (85 : 15, v/v). To remove the contaminating cholesterol and sphingolipids, which came in this fraction, we converted them to benzoyl esters and removed with a second Unisil chromatography step. The colesterol sulfate fraction obtained from this second ~hromato~aphy contained no radioactivity. Approx. 90% of cholesterol was benzoylated after 20 min reaction as described above. Higher temperature accelerated the benzoylation, but partial cleavage of the sulfate bond occurred; 17% of the cholesterol sulfate was cleav-
272
ed after 30 min heating at 80°C. The solvolysis of cholesterol sulfate, cerebroside sulfate and dehydroepiandrosterone sulfate also was completed in 4 h. The recovery of cholesterol sulfate in this procedure was examined by adding [l,2-3H]cholesterol sulfate (98 937 cpm) to 12 mg of brain total lipids. The mixture was treated as described. The radioactivity recovered in the final hydrolyzed product was 88.4%. Recovery of the added cholesterol sulfate in the final product was constant between 1 and 100 pg (Fig. 1; part of the line is not shown). The cholesterol sulfate obtained with this procedure was identified by several methods. Cholesterol sulfate was isolated from 450 g of calf adrenal by the identical procedure. Thin-layer chromatography (silica gel G as absorbent and n-propanol/concentrated ammonia/water (12 : 2 : 1, v/v as solvent) of the fraction obtained from the second Unisil column revealed a largespot with a mobility identical to that of authentic sulfate and a few minor contaminating spots. This major component was further purified by preparative thin-layer chromatography under the same conditions as described above. The band was located by spraying with bromothymol blue reagent and eluted with chloroform/ methanol/water (20 : 10 : 1, v/v). The solution was washed with water [18] until it was colorless; evaporation of the solvent yielded 5.0 mg of residue. This residue showed the same Rf value on thin-layer chromato~ams as synthetic cholesterol sulfate. The residue was converted to a sodium salt by the use of a Dowex 50.AGX-8 column [ 301. The melting point of this salt was 193-195”C,
24
11
0
Cholesteryl
Sulfate
Added
nmol
Fig. 1. Recovery of added cholesterol sulfate as free cholesterol by the present procedure. Details of the experiment were given in text.
273
while that of synthetic cholesterol sulfate sodium salt was 194-196°C: the mixed melting point was 194-197°C. These values agreed well with published data [31]. The infrared spect~m of this sulfate was identical to that of the authentic sample as previously reported f5]. The lipid portion of this sulfate was identified as cholesterol by thin-layer chromatography and gas-liquid chromatography. For the latter procedure, the material obtained from solvolysis of the cholesterol sulfate was converted to the trimethylsilyl derivative and analyzed on both OV-1 and DEGS columns. Only a single peak corresponding to cholesterol t~methylsilyl ether was obtained in both chromato~~s. Further identification of this ether was performed by gas-liquid chromatography-mass spectrometry: the molecular ion at n/e 458, as well as other characteristic fragments, such as m/e 368, 353, 329 and 255 [32] were unequivocally obtained. Concentration of ~h~~estero~ sulfate in cursors rat tissues The sulfate was present in all tissues examined (Table I). The highest concentration of the sulfate was obtained in the kidney, followed by spleen and adrenal gland. There was no significant difference due to the sex of the animal. Developmental change of cholesterol sulfate. Figs. 2A, B and C illustrate the change of tissue levels of cholesterol sulfate and of free cholesterol in rat brain,
TABLE
I
CONCENTRATIONS Data
were
presented
OF
CHOLESTEROL
as mean
Tissue
values
Tissue
obtained dry
SULFATE from
weight
IN
two
VARIOUS
separate Cholesterol dry
TISSUES
OF
43-DAY-OLD-RATS
experiments. sulfate weight)
Cholesterol (mg/g
19)
Wglg
Liver
3.72
16.40
Thymus
0.15
4.19
27.68
Lung
0.34
49.52
39.75
Heart
0.29
9.77
14.72
Spleen
0.18
77.45
33.61
249.37
20.40
Male
Kidney
0.46
Brain
0.17
20.94
Adrenal
0.01
69.38
Eye
0.03
Testis
0.27
Prostate
0.07
8.25
87.40
1.77
*
66.39
*
13.78
34.06
28.10
3.59
*
5.91
Female Liver
2.83
Thymus
0.11
4.49
24.77 41.41
20.22
8.16
Lung
0.26
56.95
Heart
0.13
9.66
15.46
Spleen
0.15
76.48
32.92
Kidney
0.37
Brain
0.72
16.24
Adrenal
0.02
50.80
Eye
0.03
2.03
*
12.92
uterus
0.05
11.51
*
31.51
Ovaries
0.02
34.31
*
52.05
* These
values
were
obtained
306.55
by pooling
tissues
from
15.11
10
71.97 *
animaIs
67.55
of the same
age,
dry
weight)
274
, /,’
x’
,r” \
1 ,L
1.\..
.
L
-ti
L
0
.
, IO
1
20
30 Age
10
50
days
Fig. 2. Change of amounts of cholesterol sulfate (open circles) and free cholesterol (solid circles) as a function of age in rat brain (A), liver (B) and kidney (C). The methods of assay were described in text. Each circle represents single determination.
275
liver, and kidney, respectively. The cholesterol sulfate level in the brain rose sharply after the 7th postnatal day, peaked at 15 days, then declined to reach the adult level at 50 days, a level which is below that observed at 7 days. This is in contrast with the free cholesterol level which increases steadily until it attains a plateau at 50 days. Liver followed an analogous developmental pattern for cholesterol sulfate; however, the peak was much broader than that of brain and the highest level was obtained at 20-25 days. The free cholesterol level in liver remained almost constant throughout the period examined. In contrast with the above two tissues, the level of cholesterol sulfate in kidney increased steadily from day 7 to day 60. After that the level plateaued. The level of free cholesterol in kidney remained constant. The rate of deposition of cholesterol sulfate in brain, liver, and kidney in lo-day-old rat was calculated as 3.75, 17.33, and 4.26 pg/g of dry tissue weight per day, respectively. Subcellular distribution of cholesterol sulfate in brain Table II shows concentrations of cholesterol sulfate in subcellular fractions of rat brain. Distributions of protein, phospholipids, cerebrosides and free cholesterol were also shown in the table, and are in good agreement with those previously reported [ 33-361. The highest concentration of cholesterol sulfate was observed in the nerve ending fractions, the level being 2.5 times higher than whole homogenate and 10 times higher than the mitochondrial fraction. The nerve ending fractions were shown to be reasonably pure by electron microscopy (Fig. 3) and marker enzyme activities: lactate dehydrogenase for nerve endings and succinate dehydrogenase for mitochondria (also shown in Table II).
TABLE
II
SUBCELLULAR
DISTRIBUTION
OF CHOLESTEROL
SULFATE
IN 3-DAY-OLD
RAT BRAIN
Values are the average of three separate experiments. See text for the method of subcellular fractionation. Mitochondria I and nerve endings I were obtained from the first gradient centrifugation and mitochondria II and nerve endings II were obtained from mitochondria I by the second gradient centrifugation. Fractions
Protein (mgl brain)
Cholester01 sulfate (nmol/ mg protein)
Homogenate Nuclei cell debris Myelin Mitochondria I Mitochondria II Nerve endings I Nerve dneings II Microsomes CYtosol n.d., not determined.
133.1 27.3 14.5 43.9 14.3 6.4 11.8 15.3 23.0
0.04 0.04 0.05 0.02 0.01 0.10 0.12 0.02 0.00
Cholester01
Cerebroside
Phosphorus
Wmoll mg protein)
Wmol/ mg protein)
WmoI/ mg protein)
0.29 0.28 0.62 0.30 0.10 0.57 0.58 0.27
